Some cells are uncontrolled in the sense that a PLMN (public land mobile network) operator does not control the location and/or operation of the cell. Other cells are controlled in the sense that the operator does control the both the location and operation of the cells. Examples of uncontrolled cells include, for example, CSG (closed subscriber group) cells, cells controlled by home enode b's and home node b's. These cells are also sometimes referred to as being “uncoordinated” in the sense that they are not subject to normal radio/cell planning.
It is generally understood that there are many methods of performing a handover for controlled cells. Inbound handover in this context refers to a cell change from a target cell's perspective where resources (for example timeslots, frequency channels, scrambling codes, etc.) for transmission and/or reception in the cell are allocated to a mobile station in advance of the mobile station performing a cell change to that cell, particularly in response to a request from the controller of the device's serving cell.
While the uncontrolled cells may be configured to use spectrum that is owned by the operator, the network operator does not have the same control over uncontrolled cells as for controlled cells. Typically, the network operator does not own the support structure (towers etc.), does not own or control the backhaul connection (e.g. Digital Subscriber Line (DSL) connections), does not know or control when a given uncontrolled cell is going to be switched on, and may not know or control the locations of uncontrolled cells. The operator will typically still have control of various parameters such as operating frequency, transmit power, etc. if the operator owns the spectrum license.
In order to perform a handover to a target cell, controlled or uncontrolled, a MS (mobile station) typically needs to provide identifying information relating to the target cell to the current serving cell so that it can initiate a handover process. In particular, the current serving cell needs to be able to communicate with (possibly via a core network) the cell controller for the target cell. However, the current serving cell may not be aware of how to reach the cell controller for the target cell, particularly if the target cell is an uncontrolled cell, unless provided with explicit identification information for the cell (such as a cell global identity). This makes handover to such cells difficult. In contrast, for a controlled target cell, it may be sufficient for the mobile station to provide the current serving cell with information about the target cell. This does not require or cause the acquisition of any broadcast information from the target controlled cell for the serving cell to be able to reach the cell controller for the target controlled cell, since the serving cell or some part of the network may be able to map other identifying features of the cell (such as operating frequency, scrambling code etc.) to the identity of the target cell or its controller.
A cell controller may not necessarily map physical layer identities (such as frequency, spreading code, etc.) of an uncontrolled cell (more generally, another cell) to the identity of its respective controller (or to an identifier which can be used elsewhere in the core network to route messages to the target controller).
An example handover procedure, including a handover preparation phase, for controlled cells is shown in FIG. 3. The PS (packet switched) Handover Command (also “PS Handover CMD” in FIG. 1) is constructed at the target cell controller based on the resources allocated during the preparation phase, and is transferred transparently via the core network and serving cell controller to the mobile station.
As part of a handover preparation, messages need to be routed from the serving cell controller to the target cell controller. In some implementations, handover is triggered in response to a measurement report, PCCN (packet cell change notification) or similar message received from the MS (mobile station). These messages identify target cells only by their physical layer identities and even this may be indirect, e.g. by means of an index into a list of such identities. These physical layer identities are not globally unique, but may be unique for a given serving cell, i.e. the tuple <target cell physical layer identity, serving cell> is typically unique for target cells that are controlled cells.
For handover to controlled cells, typically serving cell controllers are configured with sufficient information to allow messages to be routed to the appropriate target cell controller. For example, this may be by means of a mapping table of <target cell physical layer identity>→<target cell global ID>, where the target cell global ID can be used to route messages within the core network to the appropriate cell controller. Alternatively, the mapping may be stored at some node other than the serving cell controller, e.g. a core network node such as an MSC (mobile switching centre) or SGSN (serving GPRS support node).
Alternatively, the serving cell controller may be configured to know that it is also the controller of the target cell. Traditionally, this situation might be quite common, as cell controllers are organized (broadly) in geographical arrangement (i.e. many cells in a geographic locality will be controlled by the same controller). However, considering that there are now multiple different radio access technologies (with corresponding different controllers) with overlapping coverage and that new cell controllers (particularly for uncontrolled cells) are less likely to control multiple cells, this scenario becomes less common. If the cell controller is the same for both old cell and target cells, then the cell controller knows if it is able to complete the handover. The possibility of two distinct controllers adds more complexity.
If a target cell controller does not support handover (more generally if handover is not supported in respect of the target cell), the preparation phase will fail. The target cell controller may in this case indicate an appropriate cause, for example “PS Handover not Supported in Target BSS or Target system” (see 3GPP TS 48.018 sub-clause 11.3.8, covering handovers towards a GSM/GPRS/EDGE cell). To avoid repeated failed handover attempts to controlled cells, cell controllers may store information about the possibility to complete a handover towards particular controlled target cells.
Currently, cell change while in “dedicated mode” of the GERAN RAT (radio access technology), or more generally in a corresponding mode in another RAT (for example a mode during which a circuit-switched connection, such as a voice call, is ongoing) requires a preparation phase to ensure resources are available in the target cell. The preparation phase may for example include one or more of:
a request for handover sent from a serving cell controller (see for example, 3.1.5.1.1 in 3GPP TS 48.008) which is forwarded to a target cell controller (see 3GPP TS 23.009, section 6.1);
allocation of resources by the target cell controller in the target cell (for example as described in 3.1.5.2 of 3GPP TS 48.008); and
generation of a message identifying the reserved resources in the target cell (as described in, for example, 3.1.5.2.1 in 3GPP TS 48.008).